Part III-C: The Algorithm (continued)
Picking up right where we left off on the previous post...Max (name given to our data byte) is born and begins it's journey. The very second that Pat hit POST, Max entered an endless wormhole fueled by engagement. The more engagement, the faster it would travel. The first place Max arrives is the Facebook mainframe. It is here that Pat's comments, emojis, hashtags and content are defined and categorized. After applying the proper labels, Facebook then displays Max onto Pat's personal page, thus making it public viewing. At this point you may wonder what happens from here? It is a very common thought that a user's feed is synced chronologically...meaning that if a 'friend' or 'followed' user posts something, it immediately appears in their scroll feed. This is not the case.
As Max sits on Pat's user page, the Facebook algorithm is working on the backend to identify what characteristics in the content, copy and tags align with those who 'follow' or 'friend' Pat. If the content appears to be irrelevant for one of those users, the priority level of appearance drops. This is an important point, so I want to clarify this further with a small sample size.
For simplicity purposes, let's say that Pat only has 3 friends: Joe, Alan & Jenny. (As we pointed out earlier, Pat has a crush on Jenny...important information discussed in a minute). If you read the previous post, you'll understand why Pat only has three friends due to his dramatic encounters with the most basic daily actions...inside joke. So Max is live on Pat's wall, and now Facebook must decide how to make it visible to viewers of Pat's content. Each Facebook user is unique in their behaviors, tendencies, friends and other users whom they follow. Facebook's algorithm is designed to predict which of those friends and followers will entice the most frequent engagement and therefore prioritize those posts to the given user. Circling back to our sample set with Pat's friends, let's give some defining characteristics to the three amigos:
Joe:
- Has a pet dog.
- Attends the same high school as Pat.
- Plays football, basketball & baseball.
- Follows rock band fan pages.
- Engages mostly with meme posts.
Alan:
- No sign of pets.
- Has a girlfriend who has a cat.
- Attends the same high school as Pat.
- Member of the art club.
- Follows gallery pages.
- Purchases art supplies from past Facebook ad campaigns.
Jenny:
- Single.
- Attends the same high school as Pat.
- Has been tagged several times in posts with Pat.
- Comments on any post she is tagged in.
- Secretly checks out Pat's user page every now and then.
- Opens Facebook app at least 15 times per day.
Just to be clear, each of these users have a list miles long of all their activities and behaviors. Again, the idea is to simplify this for understanding how the algorithm works. As I've stated repeatedly in this series, every action you take on the Facebook platform is documented! So using Pat's pathetic friend's list, how would you anticipate Max to appear on each of their respective feeds? What priority levels would you give it? Again, in order to keep things simple, we will designate a 'high', 'medium' and 'low' priority status to each of the users on their likelihood of engaging with this type of content post.
Starting with Joe...it is clear that he has a pet, albeit a dog, but nonetheless he fits into a particular category of users. His commonalities with Pat include school and the potential of engaging with meme posts. We have no indication that Pat enjoys rock music or sports, therefore those characteristics are negligible. Conclusion: LOW
Moving on to Alan...while he does not have pets himself, Facebook has identified that his girlfriend has a cat and therefore would be interested in possibly sharing Max with his her. Again, we do not know if Pat enjoys art but we do know that he has engaged with Facebook ads in the past...therefore, align with Alan on this attribute. Conclusion: MEDIUM
Finally on to Jenny...this one should be very clear to you. Both Jenny & Pat have a crush on each other. They tag each other, 'spy' on each other's user pages, and have a tendency to engage with posts. Facebook priorities each of their content pieces to be very HIGH.
To summarize this small sample set, as soon as Max is defined by Facebook, the algorithm will position it on the feeds of Pat's followers/friends. The content, copy and engagement of the post cross-references that of the behavior patterns of the followers/friends, and then determines how to prioritize Max with all other posts flooding the system at that time. In our example, Jenny will see Pat's post the next time her Facebook feed is updated on her device. Alan and Joe will see it as well, however, Facebook allows more priority driven content to be seen first...art supply ads for Alan and rock concert dates for Joe, as an example. The priority level of visibility is important in targeting the correct users and determining their engagement behavior. We, as humans, cannot simply process that much data that quickly. The algorithm can!
.png)
